Generative "AI": The Guitar Hero of Creativity | Hacker News Reader

Generative "AI": The Guitar Hero of Creativity

whatever.scalzi.com

37 pointscheeaun70 comments

"But, like Guitar Hero, no outside creative skill has been acquired or developed."

Spoken like someone who never actually played the games. GH developed my interest in music as a kid, and Rock Band actually taught me how to drum in real life. It also garnered my interest in picking up and learning real guitar.

Those who forget that Rock Band not only a) existed and b) did everything GH did but a thousand times better + created by real musicians (Harmonix) do the genre, industry, and this article, injustice.

This article is written by someone stretching and reaching to relate two unrelated things. I'm offended by this.

Did you learn to play guitar playing guitar hero?

No, you didn't.

You're straw manning an argument that doesn't exist in the article

Not the argument you're looking for, but I've been playing drums for over ten years now because of Rock Band. Even used the sticks that came with it on my first e-kit. The game didn't teach me how to drum, but it taught me that I had a much deeper interest in playing drums than I'd ever imagined. That led into years and years of practice and I've actually become pretty good.

And that scenario is covered in the post, in the seventh paragraph. Nowhere in it is Scalzi saying Guitar Hero can’t give you an interest to pick up the Guitar, quite the opposite. The point being made is that when you do pick up the Guitar, you’re starting to learn from scratch, your years of playing Guitar Hero don’t really give you a head-start.

The paragraph in-question:

  Sure, it’s all gamified and fun and even addictive, but at the end of the day no outside, transferable skill accrues to the player. To do that, you have to set the controller down, pick up an actual guitar, and start another, entirely different journey. And perhaps some Guitar Hero player did or will, which is awesome. It means at some point they were not satisfied with pretending to play the guitar, following someone else’s rules and set-up. They still started, effectively, at zero. There was a whole new skill tree to be learned that had very little to do with the game, outside of the shape of the controller.

I get the point, but I think it's not great. A lot of what's said here can also be said about cover bands or learning to play other peoples' music.

You've had years of listening to music and practicing a very rudimentary form of sight reading.

I'm not aware of any studies proving or disproving that this gives beginners a head start, but it would be an interesting thing to investigate.

A categorical "Nice try, but you've just been wasting your time" seems overconfident.

As a person who plays several instruments (including drums) in professional situations, and as a person who enjoyed Guitar Hero, I always find it strange when folks are able to take the "experience" of Rock Band / Songstar /etc as a proxy for the experience of practicing music.

I actually do sight read for a couple of instruments- nothing in any of those games has literally anything to do with that.

Hell, my (very good) reading for clarinet and double bass has nothing to do with my (very poor) ability to sight read on piano.

Whatever gets folks interested in stuff is great, but to me it's like saying "I watched 10K hours of MTV in the 80s and that was very helpful in my career playing jazz"...

The same thing is true with writing and other skills replacements- much of my ability and desire to write comes from the experience of using it as an adjunct to thinking.

Having to figure out how I should connect ideas and state them in a way others will understand clarifies my own thoughts, so the idea that I am somehow gaining idunnowhat by adding a bunch of verbal noise to my writing misses a lot of the reasons why I like to write.

I'm happy to use LLMs to draft things like the contracts I need for the mechanical royalty payments I make when I record other folks material, but that's only because I am not trying to learn how to think legalistically.

And further, creating documents via might spark an interest in legal language and how it get used, but it's not going to help me understand the underlying reasons for what's going on any more than watching a whole lot of porn is going to help me be a better lover...

No true musician got their start in Guitar Hero, eh?

I got my start in drumming from playing on Rockband. When I mentioned it to the folk that I bought my first e-kit, they rolled their eyes. Fine, whatever, I expect musicians to be pretentious.

But yanno what? I still learned a lot. The controller itself was midi-capable so I was able to hook it up to my kit and play the game while practicing my form and timing. Was it perfect? Fuck no; it sucked. But it guided me into recognizing that I wasn't as good as I wanted to be, which led to lessons and many many hours of practicing.

I owe a lot to rock band because it opened a world of play beyond the game itself.

Nothing can invalidate whatever path you found to music.

I'm sorry you had folks rolling their eyes at you when you got your instrument. That's an exceedingly common thing to happen. Especially with singing. But it doesn't make it okay or correct or nice.

But the lessons and the practicing aren't the same thing as playing a video game.

In the same way that no matter how good I get in Fallout it will never have anything to do with how I am able to deal with folks in my real life.

For instance, I enjoy Fallout and it's very validating to play it and be the "good" guy. For a quite a while I considered that maybe there was something ethically useful in that kind of "education".

As I have gotten older I have come to feel that it's actually closer to some kind of ethics porn: none of the choices have material consequences for me or my community and I can go back and do things from a save game. I have total control over the situation in ways I never have in real life.

And further, I have come to think that this way of thinking ethics might actually be harmful to my practical ethics: because I'm not forced to make choices that actually impact me or anyone else, that game might start to erode the parts of that process that deal with the often-negative outcomes of those decisions.

So, like, sure: Fallout made me more interested in how my ethics function. But it's so far removed in how we actually move through the world that it might actually mess up how I do things.

I used to have an octapad back in the oughts, and if that was the only way I was interacting with the drum kit it would have done a lot of harm to my playing. If I played the video game drum set like I played my real kit, it would get destroyed in short order.

I am certain that you know what I mean when I say that the specific techniques around stick/beater control have virtually nothing to do with the mechanics of that kit, and further I am equally certain that you understand that the timing involved in these games has little or nothing to do with the strictness (and often looseness/feel/dragging/etc) of the timing when we play drums with other humans.

My point isn't even about the specifics of these games, but about how folks seem to abuse AI so easily these days.

I -don't- think there is anything pretentious about buying an instrument you don't play and then doing whatever you do. That's not pretense, that is the process of learning to play music- that's what it is in itself.

But I also don't think that is the same thing as playing a video game, and I think the distinction is important to make because there is much in our world that asks us to pretend we are doing things we aren't doing.

In the same way that I have to think about how to respond and think about why you have a negative reaction to what I wrote, I also have to struggle with a reasonable way of communicating why I think you don't understand what I wrote.

I have to come to terms with the limitations of my ideas and how I can state them-- if I crapped "fuck your strawman and the fleshlight version of music and thought that is generated by an LLM" into an LLM, then I wouldn't really be getting anything at all out of these kinds of interactions.

So when I read and write, I often struggle. But I think that having those difficulites and sitting with them is what is different than dropping half-baked reactions into an LLM or playing Guitar Hero.

  In the same way that I have to think about how to respond and think about why you have a negative reaction to what I wrote, I also have to struggle with a reasonable way of communicating why I think you don't understand what I wrote.

I think the point that you're missing from my post is: I don't think anybody actually says that guitar hero is like playing a guitar so I'm confused about where this idea comes from (that people think practicing guitar hero will productively lead to being able to play that). I just don't understand where that seemingly rigid argument comes from because I've not met a single person who actually believes that.

The only people who are saying that way are people who have negative feelings towards the game in the context of music. To me, that's pretentiousness because it assumes the thoughts of the people who play the game. It reminds me of people talking shit about Chicago who've never been to Chicago.

But ultimately I think my point is that people are going to flail, but we shouldn't immediately assume that people don't get anything out of them. You seem to have an absolute position that it's all or nothing -- surely you don't actually think you've learned NOTHING from playing Fallout? Nothing about yourself? Nothing about game mechanics? There's much more to learn than what's "formally" presented and each person gets something different.

Like, I've been playing with cipher mechanisms with LLMs... and after getting frustrated with all that, I'm now learning how to draw so that I can make my own similar cryptographic puzzles. Wouldn't have gotten there if I hadn't stumbled blindly for a bit. So in the same way that you wouldn't invalidate my path towards music, I'd hope you wouldn't invalidate my path towards drawing. There's a middle point somewhere in between that you seem hesitant to intellectually venture towards.

"surely you don't actually think you've learned NOTHING from playing Fallout? Nothing about yourself?"

Well, I did write out a couple of paragraphs about what I thought I learned.

I did indeed "learn how to play fallout". Which is pleasant. Like learning to play Rock Band or Sing Star or Guitar Hero was also pleasant.

And although I initially thought that maybe all those "ethically good" decisions I found gratifying might improve my ability to make ethically good choices in the larger world, I believe I was incorrect.

And further, because of the highly contrived nature of the game and it's uncanny closeness to the material world, it might be actually harmful: it encouraged me to pretend I am ethical when in fact it has brought me to do literally nothing with other people.

My point is probably getting hung up on the mechanics of drumming, which I have struggled with off and on for the last 30 years now.

So instead, I'll say what I have to say about writing and general LLM usage and be done:

The real difficulties of writing are a positive adjunct to my thinking, and I believe that almost all of the intellectual things I have struggled with (from learning to read the log files in Linux to predicting what is wrong with the engine in my car) all benefit from the same struggle.

Removing that effort by crapping half-formed thought in and out of an LLM might not be a big deal or even improve the work of folks who have already learned how to think through these complex systems.

However, that flailing certainly doesn't help people get better at thinking in the way that writing or troubleshooting complex systems might.

Further, unlike Guitar Hero an LLM might encourage folks to just "play the game" instead of actually breaking out of the bounds of that game. Because it produces tangible output (from poetry to PLC code), the folks using these "ways of thinking" often feel that they are "doing the same thing" in the same way that I can "bench press" 2-tons-of-steel with a front end loader.

And that intentional confusion between "getting better at doing stuff" and "a lot more crap got done" isn't a good thing to me.

That is the point: a lot of folks do in fact think that creating with an LLM is the same as the process of learning to create.

My point isn't to say that if your motivations for drawing or drumming or whatever come from XYZ, then they aren't valid. I'm also not claiming that picture books about airplanes are bad because they encourage children to become bomber pilots.

People have all kinds of strange and complex motivations, some better than others.

I'd trade your "enjoyment of video games" for "my parents' failures as 'professional' musicians" any day of the week.

My own reasons for going out tonight and playing double bass with my jazz band are way more sus, weird, dumb, illegitimate, whatever and have way less to do with my (very real) love of playing music with my people than anything I think some rando on the internet is going to divulge.

Rather, the point is that struggling and sucking at drawing until you figure out how to do it in a way that feels good can easily get short-circuited by midjourney or whatever.

The hidden midpoint that you think I am hesitant about might be all the pedagogical materials in-between: the scales, drum rudiments, gesture drawings, forum postings, low-rent bar gigs, half-assed bug ridden toy PHP scripts, bizarre sophomore manifestos, bad poetry, broken song lyrics, iffy recitals, lessons with bored teachers, incorrect tableture. Those things are all "pretensions" in your sense: those are all on spectra where folks flail as they approach "doing the thing". And they are all a-okay.

The limit on that spectrum, though is the LLM and the Guitar Hero, where folks are barely even "flailing". Like all parts of a spectrum are on the spectrum, but which is to be avoided if the other end of the spectrum is "good". And I probably wouldn't bother writing that out if there weren't LLMs didn't have literal cults forming around them...

> Well, I did write out a couple of paragraphs about what I thought I learned.

> And further, because of the highly contrived nature of the game and it's uncanny closeness to the material world, it might be actually harmful: it encouraged me to pretend I am ethical when in fact it has brought me to do literally nothing with other people.

I think you're underselling the second paragraph here as not learning something! "what I thought I learned" includes this lesson.

> I'd trade your "enjoyment of video games" for "my parents' failures as 'professional' musicians" any day of the week.

My dad used to be a traveling musician before he joined the army for reasons he won't tell me until I'm older (he told me that in my 30s so hmmmm), so, I totally get this.

Anywho, great chat. I appreciate the heck out of your perspective.

Jam on!

There are people here who have been using LLMs for years.

What is that?

Arrested development?

Most people who gain any sort of playing experience or value cite Rockband (for drums or Guitar mode - Rock Band 3 that shipped with a real Fender Squier) not Guitar Hero which the blog refers to. I cant understand how GH can help anyone get better at anything musical and there's literally nothing transferable to a real guitar except miming holding a plastic instrument and pressing those grotesque buttons and fundamentally different mechanical motions. You'd rather spend the same time really getting better at the instrument. You have actual music to inspire you and don't need GH to be your gateway.

The point being made isn't that it cannot inspire someone to then go on to learn but that mastery of the game alone does not convey creative skill. Of course someone can be inspired to go out and learn the skill being modeled but use of the model doesn't teach the skill even if it might make someone really excited to go learn.

Is anyone actually arguing that it conveys creative skill though?

Don’t know. Just pointing out that the comment I’m replying to didn’t actually understand the argument being made.

It seems to me TFA is arguing that it does not.

I don't know enough about DJing, but maybe that is also an analogy to GenAI usage that can work.

DJing is an interesting art. It’s not creative per se, but can be done creatively. Modern DJ software+hardware in particular allows for a ton of on-the-fly remixing, clever use of fx and stems, and even some custom on-the-fly production setups. But the average DJ (and even electronic music producer) is much closer to Guitar Hero than they are to writing their own songs for guitar.

As a code monkey generative ai helps grow my skills daily, and all I have to do is read the code.

And to keep fluffing the control narrative, 0 mention of open weight models were made.

Bad analogy. In GH you have to press the buttons at the exact time that is prescribed and you're scored on that. With AI there is no such prescribed goal. The world is your oyster. You can use it in boring ways, but you can use it in immensely creative ways too. You can chain together AI agents, image generators, video generators, 3D asset generators, computer vision models, text embeddings etc. etc. and prompt all the orchestration into existence with your ideas, taste and direction. It's the opposite of a constrained game.

Boring people use it in boring ways, interesting people use it in interesting ways.

The rest is old man yells at cloud, like how people criticized "laptop music" or photography.

This is a bad analogy by somebody who never played the game. Guitar hero is not harder or easier than guitar, it's just different.

I don't know how to play guitar, yet I can play guitar hero. My friend who plays guitar cannot play guitar hero. I'm not interested in playing guitar and I don't find GH to be a "shortcut" to learn it. In fact, I play the piano and I'm okay like that.

He's not saying playing Guitar Hero is worthless. He's saying it won't teach you guitar, which is 100% true

So, given this is all obvious, the article is basically saying nothing.

I don't know how to tell you this, but the article isn't really about Guitar Hero

Guitar Hero is a "rhythm game", perhaps calling it a "guitar simulator" is a bit incorrect. A rhythm game is one where you, for example, push buttons in tune with a song, exactly what Guitar hero is.

So you are good at different things. Maybe not at a real guitar but perhaps at reaction times.

I say this as someone who had one of those plug-in guitars but never knew how to use it, hah.

“ Being a Guitar Hero player did not — and does not! — equal being a guitar player.”

Has anyone ever claimed otherwise?

All of my childhood friends who bragged about being good at GH/RB only ever bragged about being good at the _games_. Never ran in to someone trying to pass it off as being able to play the guitar.

Anecdata, friend. Two old friends became extraordinarily good at guitar hero and even won competitions. They're also some of the best guitar players I know. One of them is an extremely good drummer and can do improv piano like a beast. They'd both agree that guitar hero helped them play better.

What are your thoughts on games like Keyboard Hero?

Maybe not, but folks keep telling me that I'm ignorant for not agreeing that being able to "engineer" a prompt is the same exact thing as being a poet or painter.

Except Guitar Hero is a game, and is marketed as such. Guitar Hero is like Bitcoin mining, you are doing useless computations to make some number go up, in the case of Bitcoin, your money, and Guitar Hero, your score.

But AI actually does something unlike Guitar Hero. There is a real output that can be used tomorrow even if AI "goes away".

You say AI helps you learn no skills, essentially. I don't know about you, but AI helped me understand Linux better in a way I didn't feel confident I would be able to before AI. I have created code with AI that I can maintain, and understand without it.

AI can go in conjunction with learning manually and can be helpful. Just because many examples exist of AI being unhelpful does not mean the whole technology is unhelpful.

I'm making a LED Camel Totem for Burning Man. I had the idea to make a Koons-style balloon animal.

I explored this idea with ChatGPT, asking it to make some reference images. I also asked for some prompt generation. I have weak sketching abilities (it's on my list to improve), so this is how I "storyboard".

To be cheeky, I asked Claude to make balloon simulator from which I could eventually drop into Fusion360 to muck with and 3D print. CamelKoons was born! [1]

I wanted a Golang backend and web frontend. Fable pretty much one-shotted the step-one scaffolding: Golang physical simulation of a single balloon, streaming geometry to a Web application. Vibed through various issues and then step two was multiple balloons with twists (which were just choke points).

In the world of disposable, hyper-focused software, this tool remained static to Dromedary Camel topologies. A generalized topology feature is left as an exercise for the reader's LLM.

I like empowering my agents to loop [2]. So, I used Claude-in-Chrome to do some loops where Claude positioned the Camel and could screenshot it. I also gave it an HTTP endpoint to position the camera for itself and me.

Eventually, I wanted to push views and comments from the Web page, so I added a comment box. Claude Code used its new-to-me "monitoring" feature to watch for that action and react to it in the session. Claude made itself a "look-at-camel" skill, so cute! [3]

While the balloons were looking interesting, I noted it wasn't very Koons-like and Claude reminded me that Koons was art and we were doing physical simulation. So then we started distorting the physical model. Of course we also added model export, etc.

After tokenmaxxing, I ended up with something that was a cross between a poop emoji and an alien dildo.

I'm now learning the Forms feature in Fusion360 and following a great balloon-dog tutorial on YouTube. I think this guitar approach will win out over my Guitar Hero approach.

[1] https://github.com/cameltopia/CamelKoons

[2] https://github.com/ConAcademy/WeaselToonCadova

[3] https://github.com/Cameltopia/CamelKoons/blob/main/.claude/s...

This seems disingenuous. I can ask a generative AI "explain this ternary operator to me" or "walk me through why this causes a race condition"

I cannot ask the video game why my real guitar is out of tune, or diagnose why I'm holding the fret wrong.

That implies that you did more than prompt the AI and actually learned whatever activity it is you are asking the AI about, because otherwise you wouldn't know to ask those questions. A person making music solely with AI won't ask about tuning their guitar because they don't have one and don't intend to use one.

One could also make hotdogs and eat them without developing any skill as a chef, but it's unlikely there would be thousands of blog posts complaining about the downsides of microwaved hotdogs every week.

If they came out with nutri-paste that gives you all of your nutrition in a grey slurry drink, and replaced everything in grocery stores with it, people would complain

And presumably there would be similarly thousands of posts saying things like "I love my nutri-paste, it lets me skip the boring part of cooking my meals straight to the eating and getting calories and vitamins which is the real purpose of eating anyways"

But nobody would be claiming that they are a chef or cooked up a novel meal, that is the point of the article.

None of that has anything to do with creativity, which is what the post is about.

Calculators have never improved my ability to do long division.

And yet, I am still quite good at mathematics.

Adorable how I'm being downvoted for stating a simple, obvious fact.

We don't appreciate facts anymore?

I was trying to figure out what time of day, by UTC, would most people be asleep. Searching I could only pre-slopped answers in hundreds of random blogs.

So I decided to get my own custom slop. Opened Gemini and it said "video generation is free for a few days, try it out!". I uploaded a copy of Neuromancer and asked it to generate a scene.

MF had Cage inserting a power plug in his ear.

Ah and the time zone curiosity question? It went about it for a a minute or so and said "can't help you with that one, anything else?"

I think we've seen the best these LLMs can do and the race to the bottom has started. It's all cost cutting now, and I am very curious to see what will be of OpenAI and Anthropic. Probably will end up absorbed by MS, Amazon and Nvidia.

You are bad at it. Don't get mad, get good.

edit: https://share.gemini.google/hvH2A3Yva5Uj

First try.

Wow, you typed out their question first try? What skill!

I doubt their attempt failed because they wrote the question poorly. It was probably just the service being flaky.

I'm bad at getting some API to not be unreliable? What are you even talking about? When your team fails an SLA do you just tell them "well it's working now so git gud"?

If you press the button on a camera, you are not a painter.

Don't worry, there's still room for paintings in human culture, even though photos outnumber them billion to one.

These folks are completely unaware that they have witnessed creative and high quality generative AI content without realizing what it was. Because that's the thing. Good generative AI is invisible per se. And it is a lot more ubiquitous than people think by now.

Like what? Can you give me an example?

> creative and high quality generative AI content

So high quality I can certainly acknowledge because indeed the output from these models can be indistinguishable from human generated content. However, I am curious to know what you mean by creative here, in two senses:

1) Do you have an example of AI generated content being particularly novel or different, and not just a (admittedly highly consistent) rehash of what it was trained on? 2) What do you think it means to describe the output of a model that doesn't possess any agency as "creative"?

I find the test of “show me AI art that isn’t a rehash” to be a bit pernicious because I have a hard time coming up with any recent examples of non-AI art that I could lie about and claim was AI generated and you’d respond “OK, that AI art is not a rehash.” If everything is a rehash, what are we arguing about?

Instead, my goto examples of high quality creative works that are clearly grounded on AI tools is https://andreaciulu.com/projects/ and https://youtu.be/fyZhC2TXgcs

“model that doesn't possess any agency” implies a model acting on its own. Like a paint brush flying around the room painting the walls with no people around. Or a computer spontaneously turning on and publishing images to an online gallery with no on operating it. The creative process of using AI tools is one of exploration, pipelining and editing.

Commentators strawman image generators as magic paintbrush machine guns. But, the process of using them intentionally is more like nature/street photography. You can’t walk into the forest and declare “I will photograph an owl fighting a fox over a mouse today.” But, you can develop a sense for where to go to find good shots. Now imagine instead of walking in a forest, you can pop anywhere in the multiverse. You don’t know exactly what to expect each time. And, the space is infinite in many dimensions. But, you can still learn to navigate.

And, that’s just the process of direct prompting. One step past that gets you configuring pipelines of highly varied generative editing tools like https://www.runcomfy.com/comfyui-workflows/comfyui-product-r...

From there you get into integrations with 2D painting/3D modeling tools and video editors. Generating references which are used to generate style sheets, then characters, sets, blockings, dialog, motion, base renders, upscales, edits.

“Make me a pretty picture. Masterpiece, best quality” is easy to make fun of. And, it’s common because so easy many people are playing around with it. But, it’s tiresome to find all of the discussion assumes that’s all there can ever be.

Scalzi has developed an almost religious sort of dogmatic hatred of AI. He seems to acknowledge none of the benefits, even those that are obvious. Likely because he feels it's harmful to the way he earns his living.

He is rapidly coming to define himself through opposition, rather than promotion. Which is a dangerous and fruitless game.

Throughout human history there has never been a still useful technology that we as a species chose to stop using, and AI very much is. Maybe we'll come to regulate it, but the genie will never go completely back into the bottle and folks are going to need to get used to the new world.

> Likely because he feels it's harmful to the way he earns his living.

Absurd. Scalzi fans are not going to stop reading his works because LLMs exist. If anything, established authors are the ones with the least to fear.

This talking point that when people dislike something it’s because it affects their personal finances is getting really tired. Not everyone is a greedy selfish prick who only thinks of themselves, people are allowed to dislike things for other reasons. That’s normal, it’s not an anomaly.

> but the genie will never go completely back into the bottle and folks are going to need to get used to the new world.

And apparently we’ll never share another original idea, either. “Genie is out of the bottle”, “you’re going to get left behind”. Apparently the people who called LLMs parrots got it wrong, they instead turn people into them, mindlessly repeating whatever the labs want them to.

Nobody even started to seriously write books using AI. But at one point there will be AI books on "most read" lists. And a more than one of those will have readership higher than anything by John Scalzi. And he's not bad, I liked what I read of his.

Plenty of people have been using AI to write books for a couple of years now.

The successful ones already knew how to write books - which directly contradicts the standard line about "It's just people who couldn't do it without AI."

That's just objectively wrong.

There have been some successful spammers who flood Amazon with tens/hundreds/thousands of titles, but that's more of an exercise in technology and marketing than writing.

Any AI bestsellers you know about yet?

You're probably right, and it will be an incredibly sad day when we have completed replaced the human part of creation in our culture with AI creation

Might just be final nail in the coffin for our cultural stagnation

Human literature won't be completely replaced, same way photography and print didn't replace painting. They just outgrew it billion times (in volmue at least) and the humanity is richer for it.

Humanity is richer because photography outgrew painting? How do you figure? I’m not judging or disagreeing, I’m just interested in the argument.

On a utilitarian sense, no doubt a photography is more useful than a painting. But in the artistic/creative sense (what the conversation is about), I’m not sure the case is that clear-cut, hence the question.

> But in the artistic/creative sense (what the conversation is about), I’m not sure the case is that clear-cut, hence the question.

Art for me is joy (or in wider definition emotion) of the beholders (and the creators). The sum of joy (and other emotions) that people derive from making, sharing and viewing photos manyfold exceeds sum of emotional value that paintings deliver.

> Apparently the people who called LLMs parrots got it wrong, they instead turn people into them, mindlessly repeating whatever the labs want them to.

Being rude about it doesn't make you right about it.

It sounds like people are repeating themselves because they have to keep repeating the obvious in every one of these threads:

The tools are useful, people are going to use them. Cope.

I use AI to code review my personal projects. There's no downside - I ignore what I don't like and take the suggestions that make sense. Most of the suggestions are good and it found a ton of little bugs over time.

I think that someone who's radically opposed to any use of AI is simply a poor developer.

There's no downside for you.

You said it yourself - you have the knowledge and ability to know what to ignore and keep what makes sense, but that's because you spent years actually writing and understanding code.

The magic wand of AI is absolutely wreaking havoc with juniors engineers' ability to understand complex code bases and software engineering fundamentals. How will the engineers of tomorrow know what to keep and what to reject if all they've ever known is "you are a principal engineer. read this Jira ticket and build a solution. ultrathink".

What you said is different from being radically opposed to AI in all situations. You're saying that junior engineers should keep learning the fundamentals of software development instead of fully relying on AI. I agree with that. Inexperienced people shouldn't be left alone with dangerous tools, so to speak, but it doesn't mean that we should abandon these tools.

I think that junior engineers can learn from AI, but the corporate setting isn't the best place for self-learning from AI. There's a lot to be learned from the average advice from the better AIs, as long as they are trying to understand the feedback. Please note that for these engineers, AI is replacing the old copy-and-paste from Stack Overflow. At least with AI they can dig deeper and ask questions.

> I use AI to code review my personal projects. There's no downside

I used to think there was no downside, but now I think there is.

You're developing the skill of choosing things out of a list to use. Your skill of creating that list is withering because you've offloaded it.

Now, I won't deny there's power in an AI code review, and for many people and occasions that tradeoff is worth it. For me, as well. But I'm convinced from my own experience it is a tradeoff.

Climate Change has accelerated rapidly. We're expecting 3 degrees of warming by 2050 now, 5 degrees by 2100. Betting on the singularity to save us is culty wankery.

I doubt that any reasonable person is actually counting on that.

It's a new, and potentially very lucrative, technology. Given time the cost will go down, not up. The way that will happen is increased efficiency.

Do I know HOW that will happen? No, but I know that it will. It always has, and there are huge motivations for making it happen.

The most reasonable and sober-minded people are counting on that:

https://www.dpg-physik.de/veroeffentlichungen/publikationen/...

I don't speak German, so I had to machine translate it... but I don't see anything about anyone counting on AI to magically solve climate change in that reference.

I've seen a few dim-bulbs or sales-types on social media say things of that sort, but never a serious adult.

Huh? Oh, I thought you were talking about the climate change part of my comment.

Everything about the data centre buildout is a bet on the singularity, that's the whole operating premise behind what all AI researchers are doing. The most grandiose charlatans of silicon valley talk about that stuff and everyone else has got in line whether or not they believe it. Trillions of dollars have spent on bringing about AGI. If it doesn't usher in AGI (and it won't) then it's all a catastrophically stupid thing to do.

Reply on news.ycombinator.com